After turning in chapter 4 last week, this week I went back and added a couple paragraphs about Schenker to Chapter 3. I also started on making nice diagrams for chapter 4, so that next week will go a bit easier.
Here’s the Mozart/Schenker, and it’s context:
3.1.5 Levels of Analysis
Visually, Functional Analysis labels can quickly guide the observer to the larger functional areas of a phrase. If an analyst wants to bring special attention to which details are more or less salient at different levels of harmony, an analysis with multiple layers of Functional Analysis is possible. This returns to the concept of prolongation. A very detailed layer can show all chords, while a more background layer can emphasize the structural harmonies that are being prolonged by the detailed layer.
The following Example 3.41 shows the theme from the first movement of Mozart’s Sonata in A major, K. 331, with three levels of analysis. The lowest level, closest to the staff, has many fine details that are often superfluous to the basic understanding of the phrase, but could be useful in some contexts for some purposes. The middle level is something like how I would normally analyze a piece, though there is still at least another level possible between the lowest and the middle. The last level shows the functional phrase pillars with lines indicating how long the Tonic prolongations last.
Typically, when I do an analysis, I start with the largest level, finding the cadences and the principal Dominants and Predominants. Then I move to that middle level, finessing the large scale analysis to note where the Predominant has a 6 instead of a 5, and which chords are embellishing the tonic. Handily, this process of adding detail very rarely involves using the eraser side of my pencil, as interpretations are more often added detail than changed roots or chords. For beginning analysis, this method of adding detail is more forgiving than methods that begin with detail before zooming out, and may help keep them from feeling frustrated.
The most zoomed-in layer provides some alternate hearings of particular chords. Measure 9 could be heard as voice-leading motion within the tonic chord or as a Predominant with the fifth in the bass embellishing tonic. Zooming into the Predominant in m. 12 shows that it is not just Predominant, but could be P with the 3rd in the bass, or it could be Pv with a 6 instead of 5, but #6. Some may also hear this as an applied Dominant, a slash-double-D with the fifth in the bass, but I prefer Pv#6 to show the transient, voice-leading nature that I perceive in this chord.
This piece could be shown with many more layers than just these three; the middle layer could be a little more or less detailed depending on the purpose of the analysis. The process of deciding how many layers to use, and determining where prolongations start and end at a larger level make the transition from Functional Analysis to Schenkerian Analysis very smooth. Deciding which pitches are important enough for the middle level of analysis, or which chords are anchoring the prolongations allows students and analysts to start thinking in a more linear fashion.
Below in Example 3.42 is a Schenker sketch of the first phrase of this Mozart Theme. Comparing the Functional Analysis levels to Schenker’s graph one can line up the similarities. Schenker’s level “c” is the surface of the music, with all the notes; his level “b” is the middleground, which takes about one bass note per beat, as we did in Example 3.41. Finally, the level “a” Urlinie shows a similar tonic prolongation as does Example 3.41, and a fruitful discussion of why certain choices were made to highlight which pitches, and why any one given interpretation is better than another for what purposes.
 Allan Keiler, “On Some Properties of Schenker’s Pitch Derivations,” 204.